G²HAN: Geometry-Guided Hierarchical AttentionNetwork for Insect Sound Classification

G²HAN: Geometry-Guided Hierarchical AttentionNetwork for Insect Sound Classification | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article G²HAN: Geometry-Guided Hierarchical AttentionNetwork for Insect Sound Classification Suyash Patil, Rekha Kaushik, Lalit Kumar This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9617709/v1 This work is licensed under a CC BY 4.0 License Status: Under Revision Version 1 posted 4 You are reading this latest preprint version Abstract Insects produce sounds through various mechanisms and organs ranging from the vibrations of wingbeats to the soundsproduced when they are feeding. These acoustic signals can be used to extract various features. This method has proven to beone of the most non invasive as well as sustainable approaches to insect classification. Traditional methods involve extractingonly one kind of features and then using Machine Learning or Deep Learning methods to classify insects. While this method canbe a great option for a very limited taxa of insects, it becomes inefficient while training a dataset consisting of many species andsubspecies. We propose a framework that aims to solve this problem by exploring the highly nuanced relationships that thesefeatures possess by learning the feature manifold and attempting to capture relational structure among numerous features.These relationships are determined by using Graph Neural Networks(GNNs). Experiments showed that these relationshipsexist and they also reveal a geometrical structure. Additionally, this work also introduces a multi-modal classification frameworkthat combines these extracted geometrical relationships in the form of embeddings as well as spectrogram features using CNN.Experimental results show that the addition of geometrical features substantially contribute to the classification performance byachieving above 97% accuracy and improving the baseline performance by over 6% Biological sciences/Computational biology and bioinformatics Physical sciences/Mathematics and computing Full Text Additional Declarations No competing interests reported. Cite Share Download PDF Status: Under Revision Version 1 posted Editorial decision: Revision requested 12 May, 2026 Editor assigned by journal 06 May, 2026 Submission checks completed at journal 06 May, 2026 First submitted to journal 05 May, 2026 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-9617709","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":635524740,"identity":"b1478ff7-936b-4238-ad13-0387664e660d","order_by":0,"name":"Suyash Patil","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABBUlEQVRIiWNgGAWjYDACZgY2hgQIE8gwsAHSjI0HiNfyoSANpKUBvxaQShiDccaHw2AWXi3m7ezPHjxgqEvccPz4s8c8Buft1rYfBtpSYxONS4vMYR5zgwSGw4kbzuSYG/MY3E7ediYRqOVYWm4DDi0SzDxsEgkMBxI3HMhhkwZpMTsA1MLYcBiPFvZnQC1Ah51//gyo5Vyy2fmHhLQwmAG1MCduuJFgJjnD4ICd2Q2CtvAAtRgcNp55442ZxAeD5ASzG0BbEvD5hf/4M8kfFXWyfefTgS78Y2dvdj794YMPNTY4tUCAAQODwgEIMxGsMgGvciiQhxpqT4ziUTAKRsEoGFkAADVLYRzf14RJAAAAAElFTkSuQmCC","orcid":"","institution":"Indian Institute of Information Technology, Bhopal","correspondingAuthor":true,"prefix":"","firstName":"Suyash","middleName":"","lastName":"Patil","suffix":""},{"id":635524741,"identity":"b21392fd-e173-4288-8df9-ab0b19c780e2","order_by":1,"name":"Rekha Kaushik","email":"","orcid":"","institution":"Indian Institute of Information Technology, Bhopal","correspondingAuthor":false,"prefix":"","firstName":"Rekha","middleName":"","lastName":"Kaushik","suffix":""},{"id":635524742,"identity":"40d89b30-5816-498d-9fb0-cdf52ce39325","order_by":2,"name":"Lalit Kumar","email":"","orcid":"","institution":"Manipal University Jaipur","correspondingAuthor":false,"prefix":"","firstName":"Lalit","middleName":"","lastName":"Kumar","suffix":""}],"badges":[],"createdAt":"2026-05-05 11:08:44","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9617709/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9617709/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":108668274,"identity":"1af8d650-1faa-4a3b-b277-bcfb58911c6b","added_by":"auto","created_at":"2026-05-07 07:06:08","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":4832784,"visible":true,"origin":"","legend":"","description":"","filename":"GGHANManuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9617709/v1_covered_1b90b500-9412-4272-a407-6b2df06250d3.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"G²HAN: Geometry-Guided Hierarchical AttentionNetwork for Insect Sound Classification","fulltext":[],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":true,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":true,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-9617709/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9617709/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"Insects produce sounds through various mechanisms and organs ranging from the vibrations of wingbeats to the soundsproduced when they are feeding. These acoustic signals can be used to extract various features. This method has proven to beone of the most non invasive as well as sustainable approaches to insect classification. Traditional methods involve extractingonly one kind of features and then using Machine Learning or Deep Learning methods to classify insects. While this method canbe a great option for a very limited taxa of insects, it becomes inefficient while training a dataset consisting of many species andsubspecies. We propose a framework that aims to solve this problem by exploring the highly nuanced relationships that thesefeatures possess by learning the feature manifold and attempting to capture relational structure among numerous features.These relationships are determined by using Graph Neural Networks(GNNs). Experiments showed that these relationshipsexist and they also reveal a geometrical structure. Additionally, this work also introduces a multi-modal classification frameworkthat combines these extracted geometrical relationships in the form of embeddings as well as spectrogram features using CNN.Experimental results show that the addition of geometrical features substantially contribute to the classification performance byachieving above 97% accuracy and improving the baseline performance by over 6%","manuscriptTitle":"G²HAN: Geometry-Guided Hierarchical AttentionNetwork for Insect Sound Classification","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-05-07 07:05:50","doi":"10.21203/rs.3.rs-9617709/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2026-05-12T12:05:41+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-05-06T12:47:39+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-05-06T12:46:58+00:00","index":"","fulltext":""},{"type":"submitted","content":"Scientific Reports","date":"2026-05-05T10:54:47+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"5d3366c5-c72e-40b6-a502-4f1a5bd28422","owner":[],"postedDate":"May 7th, 2026","published":true,"recentEditorialEvents":[{"type":"decision","content":"Revision requested","date":"2026-05-12T12:05:41+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-05-06T12:47:39+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-05-06T12:46:58+00:00","index":"","fulltext":""},{"type":"submitted","content":"Scientific Reports","date":"2026-05-05T10:54:47+00:00","index":"","fulltext":""}],"rejectedJournal":[],"revision":"","amendment":"","status":"in-revision","subjectAreas":[{"id":67639383,"name":"Biological sciences/Computational biology and bioinformatics"},{"id":67639384,"name":"Physical sciences/Mathematics and computing"}],"tags":[],"updatedAt":"2026-05-12T13:19:47+00:00","versionOfRecord":[],"versionCreatedAt":"2026-05-07 07:05:50","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9617709","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9617709","identity":"rs-9617709","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}